A Decision Tree-Based Approach for Cervical Smear

NCUTIR > College of Management > Development of Industrial Engineering and Management > 【工業工程與管理系所】期刊論文 > Item 987654321/7220

Please use this identifier to cite or link to this item: http://ir.lib.ncut.edu.tw/handle/987654321/7220

Title:	A Decision Tree-Based Approach for Cervical Smear
Authors:	楊旭豪
Contributors:	工業工程與(工程)管理系
Date:	2012-05
Issue Date:	2018-05-23 09:05:28 (UTC+8)
Abstract:	Cervical smears are used to detect cervical intraepithelial neoplasia (CIN) and remain a popular method for the early detection of precancer and cervical cancer. One of the classification systems for CIN, the Bethesda system, classifies atypical cells into ASC-US (atypical squamous cells of undetermined significance) and ASC-H (atypical squamous cells: cannot exclude a high-grade squamous intra-epithelial lesion). This paper proposes a methodology involving a decision tree (DT) to identify cases of ASC-US by constructing a classification decision tree (CDT) based on samples from a Taiwanese teaching hospital. The main difference between our methodology and those of other studies is the use of a DT to identify ASC-US cases. We calculated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) to evaluate the performance of our methodology and found a sensitivity of 86.36%, a specificity of 78.94%, a PPV of 90-47% and an NPV of 71.42%. The results indicate that our CDT is capable of detecting cases that are abnormal and positive; in contrast, detecting cases that are benign and negative remains a challenge. Moreover, our CDT performs well when predicting the cases that are abnormal and positive but performs poorly when predicting benign and negative cases. Considering both the sensitivity and PPV, we demonstrate that our CDT can help decrease the number of ASC-US cases. ©2012 ICIC International. (20 refs)
Relation:	International Journal of Innovative Computing Information and Control
Appears in Collections:	[Development of Industrial Engineering and Management] 【工業工程與管理系所】期刊論文

Files in This Item:

There are no files associated with this item.

View Licence

Loading...